Dynamic Data-Centric Storage for long-term storage in Wireless Sensor and Actor Networks

Data-Centric Storage (DCS) appears as a novel information storage and delivery mechanism for Wireless Sensor and Actor Networks in which a rendezvous node (home node) is selected to store and serve all the information of a particular application. However, DCS was not designed to provide long-term data availability. In this paper we present a Dynamic DCS solution to enable a long-term storage system. Dynamic DCS proposes to periodically change home nodes over the time based on periods of fixed duration called epochs. This makes it possible to perform temporal queries to previous home nodes in order to retrieve information from the past. We evaluate our proposal using extensive simulations, and reveal that Dynamic DCS makes sensor events available at least 85 % of the maximum lifetime provided by an optimal (but non practical) solution. Finally, we show that Dynamic DCS could easily adapt its storage performance to the requirements of an application by just tuning the epoch duration.